Authors: Jiuyun Xu; Xiao Ning; Stephan Reiff-Marganiec; Qiang Duan; Zibin Zheng
Addresses: School of Computer & Communication Engineering, China University of Petroleum, Qingdao, China ' School of Computer & Communication Engineering, China University of Petroleum, Qingdao, China ' Department of Computer Science, University of Leicester, Leicester LE1 7RH, UK ' Information Sciences and Technology Department, Pennsylvania State University, PA 16801, USA ' School of Data and Computer Science, Sun Yat-sen University, Guangdong, China
Abstract: Service composition has become a common approach to realising complex business processes. The large number of services developed and deployed independently by various providers can lead to undesirable interactions between properties of different services which are a serious obstacle for service composition to meet users' requirements. While some property conflicts can be prevented during design, many occur during execution based on runtime data. In this paper, we propose a solution for the problem of run-time resolution of service property conflicts. We formulate the conflict resolution problem as bi-objective optimisation model based on user's revenue. Solving the optimisation model provides a set of Pareto solutions which are ranked to identify the optimal one for resolving a service property conflict. The proposed scheme is implemented in a prototype for experimental performance evaluation. The experimental results indicate that our scheme is effective and efficient in resolving service property conflicts at runtime.
Keywords: web service composition; service property conflicts; bi-objective optimisation; runtime resolution; web services.
International Journal of Web and Grid Services, 2016 Vol.12 No.2, pp.142 - 161
Received: 30 Sep 2015
Accepted: 19 Dec 2015
Published online: 17 May 2016 *